Q: I'm using some 2-inch-thick lumber to make a kitchen work top. This wood has some defects, like knots. I cannot figure out how to use my hand router to cut out these defects in a curved cut and then end up with two tight-fitting pieces. Any pointers to find which way to do this would be much appreciated.

 

A: Indeed, if you cut out defects in lumber and then glue it back together, and if the cutting pattern is curved (sometimes called a serpentine cut), the joint will be nearly invisible. In fact, the U.S. Forest Service laboratory in Princeton, W.V., carried out a lot of experimental work on this technique 15 years ago. It's really amazing how this technique effectively hides the end joints.

What you're trying to do is very tough to do in one pass of a hand router. Here's why. If you carefully sketch out the path of the router when traveling on a curve, you'll see that the outside radius is different from the inside. As a result, after the router has made a curved pass and removed the defects, the two ends created will not fit together tightly; there will be a gap in the middle of the curve. If you had a numerically controlled router, you would program in the dimensions and it would cut the same radius on both ends using two passes.

If you're willing to put in the time, here's a list of steps to do the job with a hand router. After cutting out the defect with the router using a curved path, take the end of one piece and draw the approximate shape of this end on an appropriate-sized piece of stiff transparent plastic, making a pattern. Cut out this pattern and double-check to make sure it fits snugly into the first piece of wood. Attach this plastic pattern piece to the end of the second piece of wood and use it as a guide for the router. Route the end of the second piece. If this second piece fits tightly into the first piece (which it should), glue it in place.

As this involves a lot of work, I think you should consider either purchasing a numerical controlled router (tremendously useful machine) or else "farming out" this job to someone who has one.

 

Q: We bend 1-3/4-inch-thick red oak in a fairly severe bend. We have bent several hundred pieces with an 80 percent success rate. All of a sudden, we have a problem with the wood breaking. We don't steam the wood prior to bending, but use hot water. Can we hire you to come and help us out? (This is an excerpt from a long letter that describes his bending process using hot water. I do not like that process, so I will not give the details he provided.)

 

A: You can indeed hire me to help, but I think that I can solve your problem right now for free! When wood breaks in bending, there are seven items to check. Close adherence to these items will let you achieve 90 percent or higher success rates.

First, check the slope of grain in any pieces that break. Do this carefully (and safely) by putting a chisel in the end grain of the piece and then splitting the piece. The split should run down the length of the piece (that is, a split will follow the grain) and not quickly deviate to the side. If it deviates, it's because that wood has an undesired slope of grain (that is, the log was not sawn parallel to the bark). We want the grain running parallel to the sides of the piece being bent. You may have to work with your sawmillers to assure you get appropriate slope of grain in your pieces.

Second, oftentimes, we'll bend oak so that the inside and outside faces are quartersawn face rather than flat-grain faces. Did you notice the grain orientation of the pieces that broke?

Third, you need to check the incoming moisture content of the wood. This is the moisture content before you soak the wood. For severe bends, we need the wood between 22 percent to 28 percent MC. If it's wetter than 28 percent MC, then there'll be liquid water in the wood. Water is a non-compressible fluid, so when the bending takes place, hydraulic pressures will develop internally and crack the wood. If the wood is drier, it will become too brittle and too strong. Once over-dried, it's almost impossible to bring the moisture back up to an acceptable value and also restore the bendability. Specifically, soaking wood in water doesn't adequately restore the MC to over-dried pieces. Checking the incoming MC on many pieces is critical for successful bending. Note that once a piece is bent and then it fails, it's impossible to know the incoming MC unless it was measured beforehand.

Fourth, it's also important that the wood not have any surface checks that developed in drying down to 22 to 28 percent MC. Always cut samples to make sure that the wood is check-free, as any checks that are present will not be easily seen by visual inspection.

Fifth, when we bend wood, the outer radius develops tension, as it's being stretched and forced to become longer. Wood isn't always real strong in tension, so when bending we'll reduce the tension by adding some compression force to the wood in the outer radius. Most often this is done with metal straps with blocks on the end. The strap is placed over the outer radius before bending begins. Key: The wood piece must fit tightly between the blocks. When we begin to bend, the metal strap is on the outermost radius and so will automatically put pressure on the ends of the wood. This end pressure eliminates virtually all splitting. (Fairly simple, non-computerized technique that is hard to explain in words, but is standard operating procedure and it does indeed work.)

Sixth, before bending the wood, it's helpful if we heat it and also make sure that the moisture is high. Wood, when hot and wet (but no liquid water), is much more pliable. We use steam to achieve 212 F temperature and also get the moisture close to saturation without adding liquid water.

Seventh, once bent, the wood is held in the desired shape as it dries. Drying must not be too rapid or too hot. When dry, except for a slight amount of spring-back, the shape is permanent.

Have something to say? Share your thoughts with us in the comments below.